In some scenarios, such as a taxi management project use-case or context for example, a general question contemplated may be how to assist or guide a potential taxi passenger to a location where they are most likely to intercept an in-coming available taxi, in a timely and efficient manner for both the potential passenger and the taxi. In some aspects, the behavior of taxis may be learned by, for example, examining historical data related to the taxis' past performance and status, including the taxis' location, speed, direction, etc.
However, the collection of taxi status data records for even a medium-sized metropolitan city may in some instances result, on average, in the generation of 200 records/second. At this example rate, about 7 billion records can be generated in about a four month period. Furthermore, in order to address the general question posed above, the records would need to be thoroughly analyzed regarding the taxis' status, and findings of the analysis would further need to be somehow utilized.